2026-09-28 · 2026-09 / week-4

Addex Trades Below Its ATM Price, but Returned Rights Are Not Cash

Addex Trades Below Its ATM Price, but Returned Rights Are Not Cash

No-trade screen - Reject. Addex closed at $3.65 on September 25, 42.7% below the $6.37-per-ADS average at which it sold treasury ADSs in the July-August ATM. The new half-year report and the return of Indivior's GABAB portfolio make the gap look like a possible asset reset. They do not make the old financing price fair value: the returned programs remain preclinical or pre-IND, H1 operating revenue was effectively nil, and the company says its liquidity estimate still carries substantial going-concern uncertainty. The reference-only map has a flat base, a severe cash-only stress, and no evidence-backed 10% base case.

Publication time: 2026-09-28 18:19 Asia/Singapore | Scope: U.S. market, long opportunities only | Screen: Addex Therapeutics ADS (ADXN), not a selected trade Reference observation: $3.65 regular-session close on 2026-09-25, 16:00 EDT; 15,480 ADSs reported traded. This is a reference price, not an executable entry. [1]

Opportunity Ranking

Rank Candidate Direction Mispricing Evidence Freshness Catalyst Window Positioning 10/5 Status Tradeability Main Rejection Risk
1 Addex (ADXN) Long screen Quote is 42.7% below the recent $6.37 ATM sale price per ADS, but issue price is not fair value and the denominator expanded through treasury-share sales. Sep. 28 H1 report, Sep. 25 rights return and Sep. 25 quote. Management call at 10:00 EDT today; Neurosterix M4 Phase 1 completion is guided for Q4 2026. 52.97M ordinary shares sold after June; current ownership, short interest and order-book data unverified. Reject: flat base, -69.9% cash-only stress, 0:1 base reward/risk. Very low turnover; live spreads, depth and exit capacity unavailable. Rights have no disclosed license payment or funded clinical plan; additional financing can dilute holders.
2 NIO (NIO) Long screen Geely's agreements value NIO Power at RMB16B post-money, but the 30% subscription is paid into the subsidiary with both a business contribution and cash; it is not a parent cash distribution. Definitive transaction announced Sep. 28; Q2 financials dated Sep. 1. Regulatory and customary closing conditions; other commercial plans remain preliminary. No direct shareholder positioning or order-book data assessed. Reject: external unit pricing does not establish incremental parent equity value or 10/5. NYSE common/ADS; no execution audit. NIO contributes assets, receives subsidiary dilution and separately pays undisclosed cash for a 10% stake in Geely's charging subsidiary. [8]
3 Modine (MOD) Long screen Due bills begin Sep. 28 for the SpinCo distribution, but the entitlement travels with regular-way shares and the ex-spin quote is expected to adjust. Joint transaction update Sep. 17; record date is today. Expected Oct. 1 distribution and merger; final exchange ratio remains unannounced. No event-specific positioning evidence found. Reject: entitlement is not free upside; final share ratio and post-spin value remain open. NYSE common; close references differ modestly across vendors. Failed closing conditions or a changed ratio/cash distribution; the $1.90 THRM dividend is not paid on shares issued to former MOD holders. [9][10]

Selected opportunity: None qualifies. ADXN ranks first for information value, not tradeability, because today's report reconciles recent ATM proceeds, the post-period share count, H1 losses and the returned drug rights. No prior ADXN article or signal was found in the repository.

Why this one now: The same report that identifies a $6.37 recent capital-market reference also shows H1 cash of CHF0.767 million, zero contract revenue, and CHF3.456 million net loss. The issuer's claimed runway to Q4 2027 is paired with an explicit statement that substantial doubt remains. That is a new, issuer-specific cash-and-denominator bridge, not a generic biotech recovery screen. [2]

What should surprise the reader: The apparent 74.5% upside to the ATM sale price is not a sponsor bid or a valuation floor. It is the average price for company treasury ADSs sold before the rights-return announcement; the market has since traded lower while the outstanding-share count rose.

Why This Is an Information Screen, Not a Trade

The strongest bullish case is that the company has about 15 months of management-estimated runway after a $2.8 million gross ATM raise, owns a 20% interest in a private neuroscience spin-out approaching a Phase 1 milestone, and regained full rights to an Indivior-developed GABAB PAM portfolio after Indivior's parent merger. Management says one substance-use-disorder candidate completed IND-enabling work. Those are real assets and runway evidence, but the returned rights carry no disclosed upfront payment, partner, or funded clinical schedule. [2][3][4]

The countercase is stronger for the 10/5 decision. H1 revenue from customer contracts was zero; the CHF18,000 reported as income was primarily a non-cash fair-value service from Neurosterix. H1 operating cash use was CHF1.234 million; operating loss was CHF1.125 million; net loss was CHF3.456 million, including a CHF2.331 million share of Neurosterix's loss. Addex spent nothing on GABAB PAM research during H1 according to the report's program table. The company now has to choose how much of its limited capital to commit to development. [2]

Why the Price Gap Can Move More Than 5%

At $3.65, the ADS is 42.7% below the $6.37 average ATM sale price reported for July-August. A credible, funded development plan or strategic partner could prompt a move toward that past transaction price. The opposite path is also material: a call with no funded milestones, renewed partner interest, or clinical timing could leave the company financing preclinical work while its share count and cash needs grow. The issue price alone establishes neither outcome.

10/5 Asymmetry Gate

Test Reference-only result Desk threshold Finding
Highest-probability base case $3.65; 0.0% At least +10% Fails
Credible adverse stress $1.10; -69.9% No worse than -5% Fails; not a proven floor
Base reward / adverse risk 0.00:1 At least 2:1 Fails
Probability-weighted gross return -13.5% before costs Positive after costs Fails before costs
Measurement basis Sep. 25 close; reference only Verified executable entry Not met

Classification: Reject / no trade. The top case is a past capital-raise price, not intrinsic value; the bottom is a conservative cash-only sensitivity, not a recoverable value. The base is flat and the downside is far beyond 5% even before spread, slippage or future financing. Entry remains null and execution remains blocked.

The Setup

On Sep. 25 Addex said it had regained full global rights to GABAB positive allosteric modulator assets discovered with Indivior. The company attributed the return to Indivior's R&D rationalization after its parent's planned merger with Supernus. The Sep. 28 half-year report says the substance-use-disorder candidate completed IND-enabling studies and the separate GABAB chronic-cough program remains preclinical, ready for IND-enabling work. The report does not give a new licensee, funding partner, clinical-trial start date, or payment schedule. [3][4]

This differs from the information available before the Sep. 25 session: the rights return was disclosed that morning; the half-year report became public before the Sep. 28 U.S. session. The current screen adds the financial and denominator state but does not treat the partner's withdrawal as clinical validation.

The Market Price

ADXN closed at $3.65 on Sep. 25 after closing at $3.46 on Sep. 24, a 5.49% daily rise; StockAnalysis reports 15,480 ADSs traded. This is a reference close before Monday's regular session, not proof that the market accepted the H1 report. No live bid, ask, spread, depth, venue quality, volume quality or practical exit-size audit is available. [1]

The half-year report gives 203,532,333 ordinary shares outstanding at its issuance date, excluding 8,437,843 treasury shares and including 29,904,690 DSPPP shares that are treated as treasury for IFRS accounting. That reported count is 29.4% above the 157,315,524 outstanding at June 30; the report separately discloses 52.97 million shares sold after June 30, so the two figures are not interchangeable. With 120 ordinary shares per ADS, the report-date denominator implies approximately 1.696 million ADS-equivalents. At $3.65, that denominator implies about $6.19 million of equity capitalization. A vendor page reports a materially lower $3.67 million market-cap figure using 672,311 shares; I do not use it because its denominator does not reconcile to the issuer's report. [2][5][7]

The Mispricing

Observed: From July 1 through Aug. 25, Addex sold 52,970,533 shares for CHF2.3 million / $2.8 million gross. Of these, 52,762,680 were sold as ADSs through its ATM at an average $6.37 per ADS. The Sep. 25 $3.65 close was 42.7% below that average. This is a comparison to actual treasury-share sales, not a private placement or management purchase. [2]

Inference: the quote may discount the difficulty of turning returned rights into a funded clinical asset and the extra share supply. It may also be underweighting the cash runway and the possibility of a new partner. The issue-price gap is not enough to distinguish those views because the company sold shares through an ATM; the price records capital raised, not a third party's independent fair-value opinion.

The balance-sheet bridge is thin. June 30 cash and equivalents were CHF0.767 million; current liabilities were CHF1.365 million, including CHF1.263 million in payables and accruals. June 30 shareholders' equity was CHF1.909 million. Management says gross ATM proceeds raised through Aug. 25 support operations through Q4 2027, but the same report says a material uncertainty raises substantial doubt about the group's ability to continue as a going concern within a year of issuance. [2]

The Positioning

The disclosed ATM was a source of company cash and a supply of 52.97 million shares, mostly in ADS form. That supports an inference of increased tradable supply; it does not show a current short position, forced selling, or who bought the shares. The company's reported denominator and the vendor's market-cap share count conflict. Current ownership, short interest, borrow, dealer exposure, options positioning and fund flows were not established. Positioning is unknown, not bullish.

The Catalyst

Step State and timing Observable test Failure or delay path
Management half-year call Scheduled for Sep. 28, 10:00 EDT; not yet held at publication time. Addex specifies the returned candidate, development stage, budget, partner path and first funded milestone. General pipeline language without capital or a dated plan.
Neurosterix M4 PAM Phase 1 Management says completion is on track for Q4 2026. Verify enrollment/completion and any results or next-stage funding; Addex owns 20% of this private spin-out. Timeline slips or the spin-out requires more capital without a common-shareholder cash path.
Returned GABAB portfolio Rights returned Sep. 25; no new partner or clinical schedule disclosed. Verify IND/CTA plans, trial funding, molecule selection, safety package and cash requirements. Assets remain preclinical/IND-enabling and absorb cash without licensing proceeds.

The cheapest falsification test is today's management call and the next filing that reconciles cash after Aug. 25 with actual spending and the 203.53 million outstanding-share count. A new partner or funded IND would change the evidence boundary; the rights-return announcement alone does not.

The Payoff

Price Target and Probability Map

The horizon is 90 days, through Dec. 27, 2026. The scenarios are mutually exclusive subjective estimates, not observed frequencies. The top case uses the company's recent $6.37-per-ADS ATM sale price as a capital-raising reference if a concrete, funded program plan causes the market to revisit that level; it is not fair value. The base holds the latest close. The bottom case is half of an approximate gross cash-per-ADS proxy, assuming the returned programs receive no value and future spending, liabilities and fees consume half the proxy. The cash proxy itself is not a floor.

Scenario Probability Target / Level Return / Payoff Horizon Conditions Evidence Quality
Top Case 10% $6.37 +74.5% Through Dec. 27, 2026 Today's call supplies a dated, funded development plan or partner path and investors reprice toward the prior ATM sale reference. Low: no such plan or agreement is disclosed yet.
Base Case 60% $3.65 0.0% Through Dec. 27, 2026 The runway claim holds, but the clinical portfolio remains too early and no new monetization evidence emerges. Low / Medium: close is observed; path is a judgment.
Bottom Case 30% $1.10 -69.9% Through Dec. 27, 2026 Financing or program progress disappoints; market values the equity near half the gross cash-per-ADS proxy as liabilities, spending and dilution risk dominate. Low: cash proxy is stale and gross, not net asset value.
Invalidation n/a New binding partner funding or a funded clinical plan, or a further financing materially changes shares/cash. Rebuild the thesis, denominator and scenarios from the filing; no price stop is implied. At the next company filing or agreement. Primary disclosure. Medium

Probability-weighted expected value: $3.16 per ADS, or -13.5% versus $3.65, before costs. Calculation: 0.10 × $6.37 + 0.60 × $3.65 + 0.30 × $1.10 = $3.157.
Current market level and timestamp: $3.65, Sep. 25, 2026 regular-session close, 16:00 EDT.
Primary instrument: none; no trade.
10/5 favorable base move: 0.0%.
10/5 credible adverse move: 69.9% reference-only stress; not a proven bound.
10/5 measurement basis: reference-only; no executable entry.
10/5 status: Reject.
Confidence: Low. The pipeline has no current commercial revenue, positioning is unknown, and both scenario weights and the cash haircut are judgments.

Cash sensitivity: June cash CHF0.767 million converts to about $0.93 million using the Sep. 25 USD/CHF reference; adding $2.8 million of gross post-period ATM proceeds and dividing by 1.696 million ADS-equivalents gives a gross, pre-spend proxy of about $2.20 per ADS. It excludes cash consumed after Aug. 25, fees and any liquidation costs. If the bottom case used that full proxy instead of a 50% capture haircut, expected value would be about $3.49 (-4.5%), still with a flat base and an adverse move beyond 5%.

Probability sensitivity: If the top case rose from 10% to 25%, with the base held at 60% and bottom reduced to 15%, expected value would be about $3.95 (+8.1%). The highest-probability base would still be flat, the downside would still be 69.9%, and the gross base/adverse ratio would remain 0:1. A positive expected value in that sensitivity cannot rescue the failed hurdle.

The Kill Shot

The strongest counterargument is that Addex has a longer management-estimated runway after a $2.8 million raise, regains full global rights to an IND-enabling substance-use candidate without a new acquisition payment disclosed, and retains a 20% stake in a private spin-out whose M4 program may report Phase 1 completion in Q4. If today's call supplies a funded timetable, the market's 42.7% discount to the recent ATM price could narrow. [2][3][4]

The fragile assumption is that a returned asset can be advanced without costly dilution and that the market should still reference the prior ATM price. That is not established. The ATM sales were treasury-share sales, not committed capital from a strategic investor; the H1 report says GABAB PAM spending was zero during the half-year and the program was still preclinical/IND-enabling. The same report describes substantial going-concern uncertainty. [2]

What Could Go Wrong

  • The partner returned the program because its R&D priorities changed; Addex may face a higher funding burden without independent clinical validation.
  • The management runway estimate depends on assumptions and the H1 report itself says substantial doubt remains.
  • The 52.97 million-share ATM supply dilutes each existing holder's ownership proportion and the new float may keep price below the prior sale average.
  • Neurosterix losses continue; Addex's CHF2.331 million share of associate loss exceeded its CHF1.125 million H1 operating loss.
  • The cash proxy overstates recoverable common value because it uses June cash plus gross proceeds, ignores subsequent spending, and does not deduct costs of stopping programs or settling claims.
  • Thin trading can gap through any modeled level; live spread, depth, venue, volume quality and exit capacity are not known.

What Would Prove This Wrong

The rejected-long conclusion would need a fresh underwriting if Addex announces a funded partner or a board-approved clinical plan with a dated IND/CTA and complete cash budget; reports a post-August cash balance that reconciles spending and gross ATM proceeds; and provides a current denominator including treasury-share sales and plan claims. A subsequent regular-session market-quality audit is still necessary before an executable entry.

Risk Audit

Observed facts: H1 contract revenue was zero; H1 cash was CHF0.767 million; H1 net loss was CHF3.456 million; the company sold 52.97 million shares for $2.8 million gross through Aug. 25; and full GABAB rights returned on Sep. 25. Issuer estimates: operations are expected to be funded through Q4 2027 and the Neurosterix Phase 1 study is expected to complete in Q4. Model assumptions: 10%/60%/30% scenario weights and a 50% cash-proxy realization in the bottom case. These categories are not interchangeable.

Residual risks include clinical failure, IP disputes or encumbrances, future capital raises, treasury-share and incentive-plan dilution, foreign-exchange movement, going-concern pressure, reporting/control weaknesses noted in the annual report, low liquidity, trading halts, gaps, slippage and exit difficulty. No stop price would prevent discontinuous loss. No personalized recommendation is made.

Best Trade Strategy

No trade. Do not buy based on the discount to a past ATM average. Reassess after the Sep. 28 call and a filing that provides the current cash and denominator bridge, plus a funded clinical or partnering path. Before any order, verify a fresh regular-session quote, bid/ask, spread, depth, venue quality, volume quality and realistic exit capacity. No options, leverage, margin, market orders or price-floor logic are supported.

Sources

  1. ADXN historical prices: Sep. 25 close and volume, checked for this run.
  2. Addex Half-Year Interim Report 2026, issued Sep. 25 and published Sep. 28; financial statements, shares, ATM and going-concern language.
  3. Addex Sep. 28 half-year results and corporate update, sourced to Addex.
  4. Addex Sep. 25 return of GABAB PAM rights from Indivior, sourced to Addex.
  5. Addex investor stock data: one ADS represents 120 ordinary shares.
  6. USD/CHF historical rate on Sep. 25, 2026, sourced to the Swiss National Bank; used only to translate the stale CHF cash balance into a rough reference.
  7. ADXN provider market-cap history, checked against the issuer denominator; not used in valuation because the share count does not reconcile.
  8. NIO Sep. 28 Geely transaction announcement and NIO Sep. 25 historical-price context.
  9. Gentherm and Modine Sep. 17 due-bill and RMT transaction update.
  10. MOD Sep. 25 regular close, ChartExchange; finance-feed close cross-check.

Research Quality Scorecard

Criterion Score Evidence-based reason
Market disagreement 4/5 Price is well below a recent capital-raise print, but the issue price conflicts with weak reported cash flow and fresh share supply.
Evidence base 4/5 Same-day H1 report, prior ATM share-sale data, rights-return release and regular close.
Positioning and flows 2/5 ATM supply is disclosed; current holder, short, options, fund-flow and execution data are not.
Catalyst path 3/5 Management call is today and Q4 Phase 1 completion is guided, but no funded GABAB clinical milestone is disclosed.
Payoff architecture 3/5 Scenario arithmetic is reproducible; ATM price is not value and cash proxy is gross/stale.
Invalidation discipline 4/5 A funded program, current cash/denominator and later results are explicit tests.
Differentiated insight 4/5 The financing reference is reconciled across ADS ratio and share supply, and separated from fair value.
Client value 4/5 Shows why the headline discount can coexist with real dilution and going-concern risk.
Total 28/40 No-trade Reject; the score cannot override the failed 10/5 gate.

Bottom Line

ADXN's $6.37 ATM sale price is an observable capital-market reference, not an independent appraisal. The newer H1 report and returned GABAB rights strengthen the evidence file but still show no commercial revenue, a large loss, treasury-share supply and management-disclosed going-concern uncertainty. With a flat base, negative reference expected value and a cash-only downside stress well beyond 5%, the Desk rejects a long. The useful next evidence is a funded clinical plan and post-ATM cash/share reconciliation, not another comparison with the old issue price.

AI Illustration Prompt

Create an editorial illustration for The Mispricing Desk: a restrained Geneva research-finance table with two documents, one stamped “ATM SALES: $6.37 / ADS” and a second dated “SEP. 28: H1 CASH CHF0.767M.” Between them place a lab notebook labeled “GABAB: RIGHTS RETURNED, PRECLINICAL” and an unopened funding term sheet. A simple ruler marked “120 ORDINARY SHARES = 1 ADS” should show the denominator conversion. Use graphite, warm white paper, cool Swiss blue and a small amber accent; institutional, skeptical, precise, no celebratory biotech imagery. Add a subtle readable “The Mispricing Desk” watermark at lower right.